Arducopter 2.0.24b
Hopefully the long road of beta is almost over.
I've done my best to address DCM drift issues. Again, this is not the ability for the quad to hold a perfectly level spot, it's the 20-30° error introduced by vibration and rapid maneuvers.
We've upped the gain on the DCM to recover faster. There is still room to tune these gains even more aggressively. I'll be looking for input from everyone on this. To edit the gains, go into system.pde.
To optimize the hardware, please be sure to add a windscreen to your barometric sensor and solder the filter pads on your IMU shield. A piece of felt works best. No closed cell foam.
To get the best yaw performance, calibrate your ESCs, and adjust any tilted motors. The copter should NOT yaw when lifting off. If it does it's you, not me! If you want higher performance Yaw, switch from the default to rate based yaw in APM_config.h.
I added two in-field calibrations to the arming and disarming gestures. Hold disarm for ~ 10 seconds to level the copter. Hold arm for ~ 20 seconds to fly in auto-trim mode for about 45 seconds. It will exit auto-trim automatically and fly normally. Don't worry about counting out the time, just wait for the flashing lights to enter either mode.
Alt hold has been re-written and uses a single PID for both Baro and Sonar. Baro has been more heavily filtered so the D term could be brought up. It's essential for smoother alt changes. It also helps that the internal PID integrator is the same value for both now. From 3 to 6 meters, the Baro altitude is mixed with sonar for a smoother transition. The alt hold PID could use more fine tuning, let me know your experience.
This version also now includes Randy's Heli code.
Replies to This Discussion
-
Permalink Reply by Naji Karim on -
Same results here, +/-200cm sitting still on a table.
-
DeveloperPermalink Reply by jasonshort on -
There seemed to be a badly behaving filter on the temp in the BMP library. Please update.
Thanks,
Jason
-
Permalink Reply by Naji Karim on
-
Just loaded the latest version using the APM Planner.
Getting fluctuations from +50/-110 sitting still on a table. Most of the fluctuation are in the 0 and -80
-
DeveloperPermalink Reply by jasonshort on -
Sounds right. Sorry, that's the baro sensor. It's a little drifty. But then again, look what it's trying to do. Incredible.
-
Permalink Reply by vaeddyn on -
Update: Whatever got fixed, fixed my issue. Took her out for a spin this evening and things look good, alt hold under baro is give or take about a meter which I'm happy to accept compared to my previous experiences. :) She feels a lot "snappier" in general somehow, although that may be the loss of my previous PID settings.
Loiter still needs some tweaking, she oscillates a lot, but I don't think that's a code problem, I just need to get back to figuring out the right PIDs.
Thanks all, feels like 2.0.25 is in a pretty good spot.
-
Permalink Reply by Limebear (Rob) on -
Same results here with version 2.0.5 if this helps. The baro on a table fluctuates over a range of +150 to -50 ish typicaly. Take it upstairs and the reading goes up by a few metres which is right, so it's working, but very noisy. The baro is covered with headphone foam and the filter pads are soldered.
Rob
-
DistributorPermalink Reply by Dany Thivierge on -
Wow I did fly the 2.0.24Beta today!
what a nice piece of code!
I have an issue with one motor that turns too slow compared to other ones but I will just redo ESC to recalibrate tonight and try again (never had that issue before)
but even with that motor that I needed to compensate for it was (once compensated) super stable.
Thanks Jason, I did not push hard due to the motor/esc issue but tonight or tomorrow I will repeat the experience.
-
Permalink Reply by Heino R. Pull on -
Flew 2.0.24b today and things went very well. Fairly windy, but stabilize flies well - yaw holds well, and response is good. Flew alt-hold and it is very nice flying well 5 - 10 feet and holds nicely at 18 feet with wind gusts. Alt-hold pids on baro look good on my quad. I tried loiter and it is basically holding with overhoots - I need PID tuning in loiter. Flight time 8 minutes and it flies very well.
Thanks Jason and the team.
Heino
-
Permalink Reply by Gustav Kuhn on -
Heartfelt thanks Jason, for your sharing your expertise.
Bit of a storm here in Cape Town, so can only fly in my living room (small).
Loaded the new code, soldered the pads, soft foam over the baro.
Sonar appears to work well, 500mm hold over carpet.
I get gps lock inside, but doubt that 2m x 3m is big enough to test loiter !
-
Permalink Reply by Tom on
-
Jason and AC team great job on this code flys very nice i think you hit the sweet sopt really having fun flying this new code Tom
-
Permalink Reply by Naji Karim on
-
Download site showing 2.0.25 while the APM planner upload firmware is still showing 2.0.24?
Are they in sync?
-
DeveloperPermalink Reply by jasonshort on -
updated, thanks.
Social Networking
Top Content
Contests
Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.A list of all T3 contests is here.
Groups
-
ArduCopter User Group
1282 members
-
Terrain/Surface Mapping
115 members
-
ArduRover User Group
183 members
-
ArduCopter Italian Group
87 members
-
ArduPlane User Group
673 members
Advertisement
© 2013 Created by Chris Anderson.
Powered by
